The present invention relates in general to the field of file storage management, and in particular to a method for file handling in a hierarchical storage environment and a corresponding hierarchical storage environment. Still more particularly, the present invention relates to a data processing program and a computer program product for file handling in a hierarchical storage environment.
The technical area of this invention concerns the overall enhancement of file storage management. This is in an area of tremendous growth hence technologies increasing the effectiveness are very valuable.
File access typically is provided through network attached storage (NAS) systems. Tiered data storage is a technique which automatically moves data between high-cost and low-cost storage media based on rules and policies.
A request to access data that was migrated from “higher” (faster) to “lower” (slower) level storage tier typically results in a longer response time; sometimes significantly longer response time. In some tiered storage environments this might involve recalling data from removable media such as tape. In such cases it usually takes several minutes, and even up to hours, before a user can access the data if all tapes drives are currently in use. For users trying to access this data, such long response times create a bad experience, and may be simply not acceptable in certain scenarios. One example is a student in a library trying to access a series of related articles and/or files via a console that have been moved to a lower level storage tier.
Retrieving files in tiered storage environments, especially files migrated to lower level storage tiers, may present several challenges. For instance, users expect immediate file access (and thus a good user experience), however it can take a long time to actually access the data due to the slower performance of lower level storage tiers or the slow performance associated with the recall process from removable media. Moreover, users often require access to multiple files, which may be content related independent from the actual storage location or storage tier; thus the overall response time may accumulate over multiple file access. Moreover still, while hierarchical storage environment or so called storage tiering may be used because of its cost effectiveness, it may not be economic to keep all data always in the highest level storage tier. Further, conventional solutions for file access in tiered storage environments may offer limited connection between a conventionally specified set (e.g. a prior art set) of metadata associated with a file and the probability that a file will be used again.